Diabetic retinopathy is serious, costly and prevalent, especially among minority populations. Annual ophthalmic examinations and appropriate follow-up care would result in substantial savings in preventable vision loss, health care costs, and lost productivity. However, most people do not receive these services. In our previous study, we doubled the rate of ophthalmic examinations in a low-income African-American population using a multi-component intervention. We now propose to evaluate the incremental effects and costs of components of that intervention in a broader diabetes population, including Spanish-speaking patients The specific aims of the proposed study are to: 1) test the hypothesis that a tailored telephone intervention will result in a higher rate of ophthalmic examinations than a standard print intervention; 2) improve understanding about reasons why people obtain an ophthalmic examination and assess differences in subgroups (gender and Spanish language preference); 3) improve understanding about reasons why people who are diagnosed with diabetic eye disease do or do not receive recommended follow-up treatment; and 4) conduct cost-effectiveness analyses. The study design is a randomized, controlled intervention trial with masking; the individual is the unit of sampling, assignment, and analyses. A total of 800 patients with type 1 or type 2 diabetes mellitus who have not had a dilated eye examination in the last year will be sampled from the databases of two large urban health systems. After eligibility is ascertained and written informed consent is obtained, patients will be randomized within site by gender and preferred language (Spanish or English) to either the tailored telephone group or the standard print group. Sample size determinations are based on having sufficient power to detect differences between groups at the p<0.05 level of significance. The main study outcome will be receipt of a dilated eye exam at 6 months post-randomization as ascertained by medical record abstraction, with a second outcome assessment conducted at 18 months post randomization. Receipt of follow-up care for diagnosed eye disease will also be assessed. Pre- and post-intervention telephone interviews will provide data on changes in knowledge, beliefs, risk perceptions and behavior regarding ophthalmic exams. Cost data will be collected using standardized methods. Study results will inform implementation and dissemination of practical, low-cost interventions to increase ophthalmic examinations and follow-up care, and thereby contribute to a decrease in vision loss and health care costs in diabetes populations.